MC3-Project-5 - Revision history

Tucker: /* Part 2: Navigation Problem Test Cases */

2017-04-24T21:57:05Z

Part 2: Navigation Problem Test Cases

2017-04-24T21:56:41Z

Part 2: Navigation Problem Test Cases

2017-04-24T21:56:18Z

Template and Data

2017-04-24T21:56:00Z

Overview

2017-04-24T21:45:19Z

Template and Data

2017-04-24T21:45:03Z

Template and Data

2017-04-24T21:44:19Z

Template and Data

2017-04-24T21:44:04Z

Template and Data

2017-04-24T21:40:02Z

Rubric

2017-04-24T21:36:48Z

Overview

@@ Line 105: / Line 105: @@
 * T: The transition matrix can be inferred from the CSV map and the actions.
-Note that R and T are not known by or available to the learner.  The testing code <tt>testqlearner.py</tt> will test your code as follows (pseudo code):
+Note that R and T are not known by or available to the learner.  The testing code <tt>testcrlearner.py</tt> will test your code as follows (pseudo code):
 <pre>

@@ Line 66: / Line 66: @@
 ==Part 2: Navigation Problem Test Cases==
-We will test your CRLearner with a navigation problem as follows.  Note that your CRLearner does not need to be coded specially for this task.  In fact the code doesn't need to know anything about it.  The code necessary to test your learner with this navigation task is implemented in testqlearner.py for you.
+We will test your CRLearner with a navigation problem as follows.  Note that your CRLearner does not need to be coded specially for this task.  In fact the code doesn't need to know anything about it.  The code necessary to test your learner with this navigation task is implemented in testcrlearner.py for you.
 The navigation task takes place in a square grid world that measures 1.0 units by 1.0 units. The location of the robot is the "state" and it will be provided to you as a 1 by 2 ndarray where the first element represents the X location and the second element represents the Y location.  The particular environment is expressed in a CSV file of integers, where the value in each position is interpreted as follows:

@@ Line 22: / Line 22: @@
 * Update your local mc3_p5 directory using github.
 * Implement the <tt>CRLearner</tt> class in <tt>mc3_p5/CRLearner.py</tt>.
-* To test your CRLearner, run <tt>'''python testqlearner.py'''</tt> from the <tt>mc3_p5/</tt> directory.
+* To test your CRLearner, run <tt>'''python testcrlearner.py'''</tt> from the <tt>mc3_p5/</tt> directory.
 * Note that example navigation problems are provided in the <tt>mc3_p5/testworlds</tt> directory.

@@ Line 9: / Line 9: @@
 We will not review all the details of the problem here because they are available in the earlier project description.
-We will provide testqlearner.py that automates testing of your learner in the navigation problem.  Overall, your tasks for this project include:
+We will provide testcrlearner.py that automates testing of your learner in the navigation problem.  Overall, your tasks for this project include:
 * Code a continuous reinforcement learner CRLearner

@@ Line 18: / Line 18: @@
 ==Template and Data==
-The template has not yet been added to GitHub, but you can get a zip file version of it here: [[Media:Mc3p5_vrabbit.zip Mc3p5_vrabbit.zip]]
+The template has not yet been added to GitHub, but you can get a zip file version of it here: [[Media:Mc3p5_vrabbit.zip]]
 * Update your local mc3_p5 directory using github.

@@ Line 168: / Line 168: @@
 Only your CRLearner class will be tested.
 * We will create a number of groups of test cases, where each group reflects essentially the same navigation problem but with progressively higher resolution.  e.g., multiple square worlds of different sizes 5x5 world, 10x10 world, 100x100, 1000x1000, etc. Your learner will not know the dimensions of the world it is in.
 * We will test your learner against N (value of N to be determined later) test worlds with 500 iterations in each world.  One "iteration" means your robot reaches the goal one time, or the simulation times out.  Your CRLearner retains its state, and then we allow it to navigate to the goal again, over and over, 500 times.

@@ Line 5: / Line 5: @@
 ==Overview==
-In this optional project you will implement a solution the Reinforcement Learning problem with continuous multivariate state.  In an earlier project, you solved this problem using Q-Learning with one dimensional discrete states.  We will test your implementation in the context of a robot navigation problem.  The navigation problems are exactly the same as those presented in the earlier project except that instead of you being presented with a single dimensional discrete state you will receive a multidimensional continuous state.
+In this optional project you will implement a solution the Reinforcement Learning problem with continuous multivariate state.  In an earlier project, you solved this problem using Q-Learning with one dimensional discrete states.  We will test your implementation in the context of a robot navigation problem.  The form of the navigation problems are exactly the same as those presented in the earlier project except that instead of your learner being presented with a single dimensional discrete state you will receive a multidimensional continuous state.
 We will not review all the details of the problem here because they are available in the earlier project description.